Parameter Estimation for Diagnosis andOptimization in Power Plants

摘要

With the advances in the fields of controls, IT systems and software engineering, the use ofrnprocess simulation has become a crucial element for plant asset management. Clearly, thernusefulness of such systems strongly depends on the capacity of the simulation model tornrepresent the plant properties. Having accurate models on a large range of operatingrnconditions is a very challenging and largely unsolved problem. Indeed, even in cases wherernthe technical processes are well understood, as is the case in power plants, the models remainrnparameterized by a number of unknown or hardly measurable quantities. Examples arernquantities such as heat transfer coefficients in heat exchangers and pipes, component thermalrninertias, turbine and compressor thermodynamics and others.rnThis paper describes a new approach for economic optimization in power plants, based onrnparameter estimation techniques. As an example, the method is used to estimate the naturalrndegradation in gas turbine compressors and to optimize the online and offline compressorrnwashing cycles. In addition to this, the method can also be used for improving equipmentrndiagnosis and quantifying the probability of the diagnostic results. An example is given,rnshowing the capability to differentiate between different simultaneous degradation effects in the gas path of gas turbines.